Résumé

The chapter studies individual-level strategic voting in run-off elections and makes two contributions. On the theoretical side, we propose a typology of the strategic situations a voter can face and of the kind of reasoning that the rational voter should perform in such situations. On the empirical side, we conduct laboratory experiments and show that voters follow such reasoning when it is simple enough, particularly when there is a serious possibility that a candidate reaches the absolute majority threshold in the first round. More complex conditional reasonings based on ‘backward induction’ are seldom followed.